Valence, form, and content of self-talk predict sport type and level of performance

https://doi.org/10.1016/j.concog.2021.103102Get rights and content

Highlights

  • More than 85% of athletes engage in self-talk during sport.

  • Most athletes experience that self-talk improves their performance.

  • Self-talk differs across technical and endurance related sports.

  • Self-talk is qualitatively different during high- and low-pressure situations.

  • Faster marathon runners’ self-talk is less positive, short, and repetitive.

Abstract

In this paper, we aimed to test whether we could predict sport type (badminton or running) and marathon proficiency from the valence, form, and content of the athletes’ self-reported inner speech. Additionally, we wanted to assess the difference between self-talk during high intensity and low intensity exercise. The present study corroborated existing research – we were able to predict both sport type in Study 1 and intensity level as well as marathon proficiency in Study 2 from questionnaire data using machine learning models. In Study 1, we found that badminton players engage more in worry and anxiety-control while runners are more prone to task disengagement. Interestingly, it seemed in Study 2 that the more participants engaged in condensed, positive, and repetitive self-talk when not pushing themselves, the slower their fastest marathons and half marathons were. We discuss potential explanations for these findings and make suggestions for future research.

Introduction

Cognitive science classically suggests that language function is separated from other cognitive functions (Chomsky, 1986, Carruthers, 2002). The debate on the interdependence of language and cognition has been underway for millennia from Plato in the 5th century BCE over Wilhelm von Humboldt in the 19th century to most famously Whorfian linguistics in the 20th – ranging from the claim that language is merely a medium for communicating underlying thoughts between cognitive agents (Fodor, 1975), through it being a social tool for keeping track of other people’s mental states (Dunbar, 2003, Tylén et al., 2010), through language being a lingua franca between a number of specialized, quasi-modular central systems (Carruthers, 2002), to thoughts being fundamentally linguistic (Carruthers, 1996). In recent decades, many empirical studies have documented how language is closely intertwined with everything we do and modulates most aspects of behavior and cognition, including eye movements (Tanenhaus et al., 1995, Wallentin et al., 2011), perception of color (Maier and Abdel Rahman, 2018, Regier et al., 2007, Roberson et al., 2005, Winawer et al., 2007), perception of space (Levinson, 2003, Wallentin et al., 2008), respiration (MacLarnon & Hewitt, 1999), posture (Yardley, Gardner, Leadbetter, & Lavie, 1999), conditioning (Phelps et al., 2001), imagery (Stroustrup and Wallentin, 2018, Wallentin et al., 2019) and sleep (Petit, Touchette, Tremblay, Boivin, & Montplaisir, 2007). Adding to this, it is increasingly being recognized that whenever we are not engaged in overt linguistic exchange, our heads fill with inner speech and dialogue (Alderson-Day & Fernyhough, 2015). Humans spend as much as a quarter of our waking hours talking to ourselves (Heavey & Hurlburt, 2008). Self-talk is therefore gaining prominence as a topic of interest in cognitive science (e.g., Winsler, 2009, Hurlburt and Heavey, 2015, Morin et al., 2011, Carruthers, 2018, Gauker, 2018, Lœvenbruck et al., 2018).

Self-talk already plays a very prominent role in sport psychology and is endorsed by both athletes and coaches as a fruitful tool for building confidence, sustaining motivation, focusing, and improving technique (Shannon et al., 2012, Thelwell et al., 2008, Vargas-Tonsing et al., 2004, Galanis et al., 2016). It is important to study self-talk – particularly in the context of endurance and motor control, but also elsewhere – because 1) There is mounting evidence that people report to be talking to themselves (e.g. Hurlburt et al., 2013, Van Raalte et al., 2015, Gammage et al., 2001), and 2) Evidence is accumulating that talking to yourself positively affects performance in sport contexts (Tod et al., 2011, Hatzigeorgiadis et al., 2011, Van Raalte et al., 2016). In the present context, we use ‘self-talk’ to refer to covert verbalizations addressed to the self, although we recognize that a theoretically important distinction between covert and overt self-talk has not traditionally been made in the sport psychology literature. We were interested in internal self-talk as this is most compatible with existing research from general psychology on inner speech. Self-talk differs from other inner speech in that it is hypothesized to serve a specific function and to be somewhat discursive or dialogic (Van Raalte et al., 2016). Following Latinjak, Zourbanos, López-Ros, and Hatzigeorgiadis (2014), self-talk may be usefully divided into strategic self-talk often imposed by coaches and interventions and organic self-talk which is either unintentional and unbidden (spontaneous self-talk) or task-relevant and proactively used by the athlete (goal-oriented self-talk). The present study focuses on organic self-talk of both spontaneous and goal-oriented varieties.

In recent decades, the role of inner speech in cognition has received more and more attention in cognitive psychology with the empirical work mostly growing out of Vygotsky’s self-regulation theory (e.g. Vygotsky, 1986). According to the Vygotskian perspective, inner speech is the endpoint of a trajectory moving from instructions from others (parents, caregivers) over overt self-instruction (private speech) to covert self-instruction during development. The function of inner speech in this context is enhanced planning and self-control with the child gradually taking over more strategic responsibility for her own cognition and behavior. Many studies have found that inner speech also plays a role in adult cognition, facilitating a range of cognitive functions such as problem-solving, cognitive flexibility, future planning, and impulse control (Baldo et al., 2005, Emerson and Miyake, 2003, Lidstone et al., 2010, Tullett and Inzlicht, 2010). All of these cognitive skills are needed to perform well in sports (see parts IV ‘Motivation and Emotion’ and V ‘Cognition’ on pp. 313–449 in Handbook of Sports Psychology: Schinke et al., 2016), especially in situations that are challenging in terms of technique or duration (Theodorakis et al., 2000, McCormick and Hatzigeorgiadis, 2019). Tod et al. (2011) preface their systematic review on self-talk in sport by discussing evidence from the literature indicating that novice athletes may engage more in and benefit more from self-talk than more experienced athletes (compare young children and adults’ self-directed, self-regulatory speech). Similarly, Hatzigeorgiadis et al. (2011) found that self-talk was more effective for novel tasks than familiar tasks, again supporting the idea that the inner voice is used for self-regulation in challenging situations. Recently, sport-specific theoretical models of self-talk have been proposed in response to the increase in empirical investigations – the theoretical debate has taken place in particular between the goal-oriented versus spontaneous self-talk framework first proposed by Latinjak, Zourbanos, Lopez-Ross, and Hatzigeorgiadis (2014) and a model based on System 1 and System 2 framework as well as a discursive perspective on inner speech (proposed by Van Raalte et al., 2016).

Latinjak et al. (2014) argued that self-talk may be usefully conceptualized in a dichotomy of goal-directed and undirected thought processes, thus modifying an original framework from cognitive neuroscience by Christoff (2012) – here, goal-directed thoughts usually occur in contexts requiring reasoning, problem-solving, and decision-making while undirected thoughts encompass thoughts or self-talk statements that come to mind effortlessly and automatically and are not necessarily related to the task or context at hand. Goal-directed thought/self-talk is deliberately put to use when solving a task, akin to when children talk to themselves out loud to direct attention and control action. It is used to represent the current model of the self and the desired model of the self and prompts actions to convert the current model to the desired model (Unterrainer & Owen, 2006). Christoff’s original framework is motivated by the fact that goal-directed thinking has been disproportionately emphasized in cognitive neuroscience, often at the expense of undirected thinking which is difficult to elicit in experimental settings. Instead, she argues that we should expand our focus to include undirected thinking which has traditionally been conceptualized as a kind of drift without a function. Latinjak and colleagues adapt this framework to sport psychology research and argue that we should also investigate self-talk as either goal-directed or undirected. While Latinjak et al.’s adapted framework has been influential and many studies have adopted it, it has also received criticism (see especially Van Raalte, Vincent, Dickens, & Brewer, 2019), mainly for being too focused on theoretical taxonomy and neglecting to address the validity issues surrounding questionnaire-based approaches. These critics also point out that the framework does not make immediately testable predictions.

Although Van Raalte and colleagues have criticized Latinjak et al.’s framework, their own is in fact similar in many ways. Van Raalte et al. (2016)’s model aims to take into account the discursive nature of self-talk in light of dual-process theories (Evans & Over, 1996; Sloman, 1996, Kahneman, 2011, Stanovich and West, 2000). This model is composed of two perspectives: A) self-talk is discursive in nature, can occur internally or externally, has a function, is ‘syntactically recognisable’, and B) self-talk can be divided according to dual-process theories where one system is fast, intuitive, shallow, and the other system is slow, elaborate, complex. The ‘discursive nature’ of self-talk refers to the idea that inner speech is essentially a dialogue between many different internalized positions. These different positions are hypothesized to be the individual athlete’s beliefs, bodily reactions, and experiences which he or she represents and reacts to using language. Within this kind of model, System 1 and System 2 self-talk are hypothesized to serve different functions. System 1 represents the immediate, emotionally charged reaction to a situation, such as when athletes automatically swear at themselves when they make a mistake. System 2 represents top-down, strategic use of self-talk whether this takes the form of motivational, instructional, or distracting self-talk. System 2 also monitors System 1 and attempts to regulate the automatic output. The dual-process perspective and the discursive perspective are compatible – according to Van Raalte et al. (2016) – because they together explain how we use self-talk (by communicating between different “internalized positions”) and why (to use System 2 to control System 1 as well as our behavior).

The discursive aspect of Van Raalte et al.’s model is compatible with a Vygotskian model of self-regulation (see especially Larrain & Haye, 2012) as Vygotskian self-regulation is also hypothesized to be internalized dialogue stemming from overt dialogue with parents or care-givers. Similarly, the dual-process perspective is compatible if we think of System 2 as the self-regulating system, reacting to the automatic inner speech emerging from System 1. As System 2 is supposedly quasi-rational and amenable to further information coming from the outside, we should also envision it as the system that actually undergoes change during self-talk interventions (this would be termed ‘strategic self-talk’ in Latinjak et al.’s terminology). This happens for example when athletes during a self-talk intervention learn to override their negative thoughts in response to defeat and focus their inner voice on technique instead.

The matching hypothesis (Theodorakis et al., 2000) proposes that self-talk differs depending on the type of sport. ‘Matching’ in this case refers to the idea that self-talk should ‘match’ the task at hand in terms of form and content in order to be effective. According to the original matching hypothesis, instructional self-talk should be more effective than motivational self-talk for sports depending on fine motor skills requiring precision and timing (such as dart throwing) while the reverse should hold for sports involving gross motor skills requiring strength and endurance (such as cycling). The evidence in favor of this original matching hypothesis is mixed. The systematic review by Tod et al. (2011) did not find support for the matching hypothesis as they found no difference between the effects of instructional and motivational self-talk on performance on gross and fine motor skills tasks. On the other hand, a meta-analysis by Hatzigeorgiadis et al. (2011) did find support for the matching hypothesis as instructional self-talk interventions were more effective for fine motor tasks than motivational self-talk interventions were, and instructional self-talk was also found to be more effective for fine motor tasks than for gross motor tasks. More recent developments of and additions to the matching hypothesis have refined its predictions, for example suggesting that instructional self-talk may be more effective at early stages of learning (see e.g. Zourbanos, Hatzigeorgiadis, Bardas, & Theodorakis, 2013). Indeed, Hatzigeorgiadis et al. (2011) found that task familiarity moderated the effect of self-talk on performance with the effect for novel tasks being on average larger (d = 0.73) than the effect on learned tasks (d = 0.41). Conceptualizations of the function of self-talk in sport have, as indicated above, undergone drastic developments over the last decade (see Latinjak, Hatzigeorgiadis, Comoutos, & Hardy, 2019, for a summary), and one further aspect that has been refined is the idea that different skill-levels and different types of sport place different demands on self-talk (see e.g. Hatzigeorgiadis, Zourbanos, Mpoumpaki, & Theodorakis, 2009). Thus, we were interested in how organic self-talk (i.e. not imposed from interventions) differed between different sports.

Only a few studies to date have directly investigated self-talk in endurance sport. This type of sport is particularly interesting from a cognitive perspective because it involves a real-world example of a challenge that requires a sustained high degree of cognitive control. Long-distance runners, cyclists, swimmers, rowers, etc. have to continuously inhibit the prepotent response (slowing down/quitting) in order to fulfill a longer-term goal (winning/completing). Performing this type of activity also offers rich opportunities for self-talk as the athletes are often alone with their thoughts for prolonged stretches of time during both training and competition. The self-talk content in endurance sport has previously been investigated in a questionnaire study on marathon runners by Van Raalte et al. (2015) and seven intervention studies (Blanchfield et al., 2014, Wallace et al., 2017, Hatzigeorgiadis et al., 2018, Hamilton et al., 2007, Schüler and Langens, 2007, McCormick et al., 2018, Barwood et al., 2015). Van Raalte et al. (2015) asked a large number of marathon runners (N = 483) to list their thoughts in an open-ended way and later coded the answers into categories. The authors found that 88% of marathon runners reported engaging in self-talk while running, and that this self-talk took a variety of both motivational and instructional forms. Although their study did not explicitly associate self-talk with marathon performance, Van Raalte and colleagues did find that elite marathon runners meeting standards set by USA Track & Field engaged significantly more in associative self-talk (i.e. self-talk related to the body and running technique) than non-elite marathon runners.

The intervention studies investigating endurance sport vary widely in type and duration of intervention, the inclusion of a plausible, active control group, and the kind of physical activity investigated. For example, Blanchfield et al. (2014) found that a two-week motivational self-talk intervention was associated with both improved endurance performance on an exercise bike and reduced ratings of perceived exertion compared to a passive control group. The remaining laboratory-based studies found that positive (Hamilton et al., 2007) and motivational (Barwood et al., 2015) self-talk improved endurance cycling performance, including under conditions of uncomfortable heat (Wallace et al., 2017, Hatzigeorgiadis et al., 2018). Specifically, with regard to endurance sports, it is worth noting that only two intervention studies to date have taken place outside the laboratory, thus limiting the ecological validity of the findings. In these two “field studies”, one found an effect of self-talk on endurance sport (Schüler & Langens, 2007), while the other did not (McCormick et al., 2018). McCormick et al. (2018) found that runners who were trained on motivational self-talk reported finding the intervention helpful although it did not affect their performance on a 60-mile, overnight ultramarathon. Schüler and Langens (2007) found that self-talk training provided a successful buffer against the negative impact of psychological crises during a marathon race.

The present study aimed to replicate existing findings on the prevalence and efficacy of self-talk in sports and relate these findings to cognitive science research more broadly. In addition, its goal was to explore the effects of personal and contextual factors on self-talk (Van Raalte et al., 2016) with Study 1 comparing self-talk across sport contexts (type of sport) and Study 2 exploring how personal factors (e.g., skill-level) are related to self-talk and performance.

It is often assumed in the literature that different types of sports place different demands on self-talk (see for example the matching hypothesis; Theodorakis et al., 2000), but direct comparisons of different sports, e.g. involving different levels of fine motor control, have rarely been conducted in the same study. In Study 1, we tested whether it was possible to discern the types of self-talk produced by practitioners of a sport involving a high degree of technical control (in casu badminton) versus a sport relying more on endurance (running). These two types of sport are also notably different in that badminton is a game while running is often a timed event. In Study 2, which dealt exclusively with long distance runners, we also tested whether self-talk was associated with marathon/half marathon running ability as well as whether participants differed in their self-talk use in high pressure versus low pressure situations. We hypothesized that high effort would be associated with abbreviated, associative self-talk and that low effort would be associated with dissociative self-talk in fuller sentences. With regard to the connection between self-talk and skill-level, there is evidence that a lower skill-level is associated with more self-talk, both in children learning a skill for the first time (e.g. Kray, Eber, & Karbach, 2008) and in adults performing difficult tasks (e.g. Emerson & Miyake, 2003). In Study 2, we therefore investigated the relationship between self-talk and skill-level as well as differences in self-talk associated with high and low effort.

In both Study 1 and Study 2, we also asked participants how often they exercised, as we hypothesized that people who exercise frequently will be more adept at using self-talk as a tool for motor control and endurance.

Section snippets

Overview

The Automatic Self-Talk Questionnaire for Sports (ASTQS; Zourbanos, Hatzigeorgiadis, Chroni, Theodorakis, & Papaioannou, 2009) is a questionnaire made to measure the quantity and quality of self-talk used by athletes of varying levels of activity and fitness. The questionnaire measures four positive and four negative self-talk dimensions. Positive self-talk consists of psych-up (e.g. ‘come on’), confidence (e.g. ‘I’m very well prepared’), anxiety-control (e.g. ‘don’t get upset’), and

Questionnaire reliability

We calculated the internal consistency (Cronbach’s alpha) of the entire questionnaire using the following formulaα=pp-11-Σi=1pσyi2σx2where p is the number of items, σx2 is the variance of the observed total test scores, and σyi2 is the variance of the ith item. The questionnaire had a Cronbach’s alpha of 0.889, indicating a good internal consistency.

Self-talk frequency and self-talk efficacy

See Table 1 for an overview of answers split into runners and badminton players and Fig. 1 for visualizations of the responses.

Sport type

We fitted three

Overview

For Study 2, we wanted to explore how self-talk differed in terms of form and content in high intensity versus low intensity situations. Our new questionnaire queried participants’ self-talk in high versus low effort situations (in both training and competition) and asked about positive content, negative content, other-athletes focused content, and task-irrelevant thoughts. In terms of the form of self-talk, the questionnaire asked about length and variety. We did not find this combination of

Questionnaire reliability

The questionnaire we used in Study 2 only had a Cronbach’s alpha of 0.601 (see Study 1 for how it was calculated), indicating poor internal consistency, i.e. that the items are unlikely to measure only one underlying concept. This is not surprising given that we designed the questionnaire to measure self-talk under high and low pressure conditions and thus the questionnaire should not be unidimensional. There were not enough items to calculate Cronbach’s alpha for high and low pressure

Discussion

Our results from both Study 1 and Study 2 corroborated results from the existing literature suggesting that athletes talk to themselves while exercising and that they generally believe it helps them perform better. Using machine learning models, we also found some indication that you can predict both people’s skill-level and type of sport from the content of their self-talk. This provides support for the idea that self-talk actually matters and is not just an epiphenomenon of mental life. Our

Conclusion

The present study replicated existing findings on the content, frequency, and function of self-talk in sports. We were able to use answers to a self-talk use questionnaire to predict specific sport (running versus badminton), pressure level (high versus low), and marathon running proficiency. Our findings from Study 1 suggest that self-talk in an individual, gross motor sport like running can be characterized by more disengagement compared to a team-based, more technically demanding sport like

Funding

This work was supported by the Department of Linguistics, Cognitive Science and Semiotics at Aarhus University.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors would like to thank Nikos Comoutos (former Zourbanos) for access to the Automatic Self-Talk Questionnaire for Sports and the anonymous reviewers for insightful comments on earlier versions of this manuscript.

References (80)

  • J.S. Lidstone et al.

    The roles of private speech and inner speech in planning during middle childhood: Evidence from a dual task paradigm

    Journal of Experimental Child Psychology

    (2010)
  • A. Morin et al.

    Self-reported frequency, content, and functions of inner speech

    Procedia-Social and Behavioral Sciences

    (2011)
  • D. Roberson et al.

    Color categories: Evidence for the cultural relativity hypothesis

    Cognitive psychology

    (2005)
  • J. Thibodeaux et al.

    What do youth tennis athletes say to themselves? Observed and self-reported self-talk on the court

    Psychology of Sport and Exercise

    (2018)
  • A.M. Tullett et al.

    The voice of self-control: Blocking the inner voice increases impulsive responding

    Acta Psychologica

    (2010)
  • J.M. Unterrainer et al.

    Planning and problem solving: From neuropsychology to functional neuroimaging

    Journal of Physiology, Paris

    (2006)
  • J.L. Van Raalte et al.

    Self-talk: Review and sport-specific model

    Psychology of Sport and Exercise

    (2016)
  • M. Wallentin et al.

    Eye movement suppression interferes with construction of object-centered spatial reference frames in working memory

    Brain and cognition

    (2011)
  • N. Zourbanos et al.

    The social side of self-talk: Relationships between perceptions of support received from the coach and athletes’ self talk

    Psychology of Sport and Exercise

    (2011)
  • B. Alderson-Day et al.

    Inner speech: Development, cognitive functions, phenomenology, and neurobiology

    Psychological Bulletin

    (2015)
  • Barwood, M.J., Corbett, J., Wagstaff, C.R.D., McVeigh, D., & Thelwell, R.C. 487 (2015). Im- provement of 10-km...
  • A.W. Blanchfield et al.

    Talking yourself out of exhaustion: The effects of self-talk on endurance performance

    Medicine and Science in Sport and Exercise

    (2014)
  • Brinthaupt, T. M., Boling, J. S., & Wilson, C. (2000). Situational and individual differences in the self-talk of...
  • P. Carruthers

    Language, Thought and Consciousness: An Essay in Philosophical Psychology

    (1996)
  • P. Carruthers

    The cognitive functions of language

    Behavioral and Brain Sciences

    (2002)
  • P. Carruthers

    The causes and contents of inner speech

  • N. Chomsky

    Knowledge of language: Its nature, origin, and use

    (1986)
  • R.I. Dunbar

    The social brain: Mind, language, and society in evolutionary perspective

    Annual Review of Anthropology

    (2003)
  • K.A. Ericsson et al.

    Verbal reports as data

    Psychological Review

    (1980)
  • J.S.B. Evans et al.

    Rationality and reasoning

    (1996)
  • J.A. Fodor

    The Language of Thought

    (1975)
  • J. Friedman et al.

    Regularization paths for generalized linear mod-els via coordinate descent

    Journal of statistical software

    (2010)
  • C. Gauker

    Inner Speech as the Internalization of Outer Speech

  • R.A. Hamilton et al.

    Assessing the effectiveness of self-talk interventions on endurance performance

    Journal of Applied Sport Psychology

    (2007)
  • J. Hardy et al.

    A framework for the study and application of self-talk within sport

  • J. Hardy et al.

    Quantifying athlete self-talk

    Journal of Sports Sciences

    (2005)
  • A. Hatzigeorgiadis et al.

    Self-talk and sports performance: A meta-analysis

    Perspectives on Psychological Science

    (2011)
  • A. Hatzigeorgiadis et al.

    Beat the heat: Effects of a motivational self-talk intervention on endurance performance

    Journal of Applied Sport Psychology

    (2018)
  • D. Kahneman

    Thinking, fast and slow

    (2011)
  • J. Kray et al.

    Verbal self-instructions in task switching: A compensatory tool for action-control deficits in childhood and old age?

    Developmental science

    (2008)
  • Cited by (6)

    View full text